Few studies to date have addressed the role of the descending system in joint pain, except for a number of electrophysiological studies of descending inhibition, and only in the acute phase of the pain state [39–41]. Here we have combined lesion of the RVM with behavioural studies to address the role of descending facilitation in a model of inflammatory joint pain. In addition microarray analysis has led to the identification of a number of possible mediators of descending facilitation at the level of the dorsal horn.
We first investigated the role of the descending 5-HT pathway in the development of mechanical hypersensitivity following ankle injection of CFA. Previous studies have indicated the descending 5-HT pathway contributes to the maintenance of behavioural hypersensitivity in neuropathic pain . In the present study intrathecal administration of the selective toxin 5,7-DHT attenuated mechanical hypersensitivity at 1d and 2d post CFA injection, suggesting that the 5-HT pathway contributes in a time-dependent manner to hypersensitivity in this pain state. In the hours following inflammation and at later time points, the 5-HT system does not appear to facilitate the pain state. This supports previous findings by others that siRNA silencing of the tryptophan hydroxylase (TPH) enzyme within the RVM attenuates mechanical hypersensitivity at 1d – 3d following plantar injection of CFA .
Our second lesion experiment targeted the electrophysiologically defined ON cells of the RVM, which are believed to be facilitatory in nature . Although ON cells do not have a single neurochemical classification, they can be identified by their direct responsiveness to morphine implying that they express the mu opioid receptor (MOR) [19, 20]. This characteristic has been exploited previously by others to ablate the MOR expressing (MOR+) neurons of the RVM using the selective neurotoxin dermorphin-saporin [21–23]. We found that, as with 5,7-DHT treatment, MOR+ cell depletion attenuated mechanical hypersensitivity. However this effect was more prolonged than for 5,7-DHT depletion, with significant attenuation observed from 1d to 7d post CFA injection. This increased attenuation may have been because some MOR positive neurons are also 5-HT positive  and lesions of RVM therefore would have caused partial ablation of both descending control pathways.
However while both the 5-HT and MOR+ cell pathways played a role in regulating the mechanical hypersensitivity associated with joint inflammation, the magnitude of attenuation in both lesion studies was considerably smaller than that observed in neuropathic pain states [21–23]. Similarly the time of onset of descending facilitation differs in the present study than that shown previously for neuropathic pain. The previous dermorphin-saporin studies of neuropathic pain indicated that descending facilitation is required for the maintenance but not the induction of behavioural hypersensitivity [21–23]. In contrast, descending facilitation of joint pain is evident from one day onwards. This suggests that in neuropathic injury, descending facilitation is required for the maintenance of the pain state, whereas in joint inflammation descending facilitation contributes to both the induction and maintenance phases. It should be noted that we have measured mechanical hypersensitivity of the hindpaw, which is a reflection of secondary hyperalgesia. We did not address primary hyperalgesia of the inflamed joint, which would require measurement of load bearing and other more direct measures of primary sensitization. Nonetheless, the present findings support previous studies which suggested that descending facilitation modulates secondary hyperalgesia [43–45].
While the contribution of descending facilitation to central sensitisation and behavioural hypersensitivity in a variety of pain states has been extensively studied [21, 22, 24, 25] little is known about the molecular mechanisms underlying descending facilitation. Microarray analysis has however been used to characterise changes in dorsal horn gene expression at 2 h-7d following ankle injection of CFA . To build on these findings, the present study aimed to examine the contribution of descending facilitation to the regulation of gene expression at the 7d time point. Microarray analysis was carried out on two groups, both with CFA inflammation, and with dermorphin-saporin or saline pretreatment as the only variable. In this way the analysis revealed genes that are regulated in this pain state by the MOR+ RVM descending pathway.
A number of cellular mechanisms could potentially contribute to descending facilitation of dorsal horn excitability. Changes in receptor availability, presynaptic release of neurotransmitters, post-synaptic changes in intracellular signalling pathways, transcriptional regulation and immune cell activation are all key features of central sensitisation  and could be subject to modulation by descending facilitation. However, surprisingly few genes with neuronal annotations were identified in our analysis, while many genes associated with immune cell function were identified.
A prominent role for immune cells in dorsal horn in mediating central sensitisation and behavioural hypersensitivity has been established [16, 17, 47, 48] and immune system related genes are a feature of many other microarray studies of the dorsal horn [47–49]. Many of the immune related genes identified by our microarray analysis were found to be downregulated by dermorphin-saporin treatment, suggesting that descending facilitation by the MOR+ cell pathway is a positive modulator of immune cell processes within the dorsal horn.
Among the genes confirmed by RT-qPCR was Nos2, which encodes the enzyme inducible nitric oxide synthase (iNOS). Nos2 gene expression is induced in peripheral immune cells during inflammation, and has also been shown to be regulated in microglia and astrocytes . Nitric oxide (NO) itself has been implicated in central sensitisation in various pain states [32, 51–55]. Although the focus has largely been on neuronal NOS as the enzyme responsible for spinal production of NO, a number of reports have implicated iNOS protein in inflammatory and neuropathic pain states [32–35]. Bilateral induction of Nos2 in the dorsal horn has been shown to occur following ankle injection of CFA  and our present data suggests that Nos2 expression may be modulated, at least in part, by descending pathways. Interestingly upregulation of Nos2 occurs in the spinal cord in a model of stress-induced hyperalgesia, and CCK signalling within the RVM is required for this process . Our finding also suggests that Nos2 gene expression in the dorsal horn is modulated by descending pathways.
One of the key findings of our present study was the downregulation of two members of the same family of chemokines in the dermorphin-saporin treated group. The expression of these genes was not altered by CFA joint inflammation alone . Chemokines are small molecular weight proteins which regulate leukocyte migration and activation and act as a key component of the immune response . Within the dorsal horn, a number of chemokines have been shown to play a role in various pain models and there is an established role for chemokines such as CX3CL1 (fractalkine) and CCL2 within the dorsal horn in both neuropathic and inflammatory pain states . Cxcl9 and Cxcl10 are members of the interferon gamma (IFN-γ) inducible subset of CXC chemokines. A number of previous studies have suggested a role for the cytokine IFN-y within the dorsal horn in chronic pain states [58–61]. As targets of IFN-y signalling, Cxcl9 and Cxcl10 may therefore be involved in nociception in the dorsal horn. Notably the shared receptor for these chemokines, Cxcr3, was also downregulated in the dermorphin-saporin treated group. The simultaneous downregulation of these Cxcl9, Cxcl10 and Cxcr3 suggests that this chemokine-receptor signalling pathway may contribute to descending facilitation via the MOR+ cell pathway, indicating a novel role for this chemokine family in pain signalling within the dorsal horn.
Although most studies of CXCL10 in the CNS have focused on its expression in immune cells and role in neuroinflammation  one study has demonstrated that CXCL10 is expressed constitutively in neurons in culture, and may be released tonically at low levels . Furthermore exposure of hippocampal neurons in vitro to exogenous CXCL10 causes increased excitability . Although to date, chemokines within the CNS have been investigated predominantly as modulators of immune cells, in future it may be of interest to investigate their roles in neuronal transmission .
The expression of CXCL9 and CXCL10 is also increased in the joints of patients with rheumatoid arthritis [65, 66] suggesting that these chemokines may contribute to the joint pathology associated with the disease. Indeed CXCR3 receptor antagonists have also been shown to be effective in decreasing the pathogenesis of rheumatoid arthritis in an animal model . A pronociceptive role for CXCL9, CXCL10 and CXCR3 signalling within the dorsal horn would provide further support for the use of chemokine antagonists in the systemic treatment of joint pain states.
In conclusion, the present data demonstrates for the first that time that descending facilitation contributes to behavioural hypersensitivity following joint inflammation. This implies that in addition to targeting the underlying joint pathology, patients suffering from joint pain symptoms may benefit from treatments that reduce central sensitisation and descending facilitation. In particular, the CXC family of chemokines are subject to regulation in the dorsal horn by descending facilitation and therefore may be promising targets for further studies.